The High Resolution Estimate-a Comparative Study

Shahid Niaz,Fang Dagang,Sheng Wei Xing,Ye Xiao Dong
DOI: https://doi.org/10.1109/iccea.1999.825121
1999-01-01
Abstract:High resolution signal parameter estimation is a significant problem in a wide range of applications particularly for radar imaging and computational electromagnetics. In this paper, we present a comprehensive study of a few notable high resolution techniques for parameter estimation like MUSIC, matrix pencil and RELAX based upon numerical simulation and compare their estimation performance. The result shows that RELAX provides better estimates for closely spaced signals especially for low signal-to-noise ratio (SNR).
What problem does this paper attempt to address?